Characterization of diversity at the killer immunoglobulin-like receptors (KIR) locus has progressed over the past several years, particularly since the entire sequence of the complex became available. To date, segregation analysis has been used to estimate full KIR haplotypes in four studies, representing a total of 79 families (roughly 316 independent KIR haplotypes). Based on the previously published gene order, we have compiled a listing of 37 unique haplotypes derived from these studies. Although it is possible that one or more of these haplotypes are incorrect, as a number of assumptions had to be made when deducing the haplotypes, it represents an astonishing number of haplotypes given the relatively few independent chromosomes that were studied. In order to define more precisely the extent of haplotypic variability based on gene content among individuals of northern European descent, we have completed typing of 59 Centre d'Etude Polymorphisme Humain (CEPH) families (> 400 independent chromosomes) for presence/absence of all KIR genes. KIR2DL4 has been subtyped by sequence analysis, and minimal subtyping of other KIR genes has been or will be performed prior to thorough statistical analysis of the haplotypes. Clearly, expansion and contraction of the haplotypes have occurred (in some cases to an extreme degree), perhaps due to unequal crossing over in the region. The CEPH KIR haplotype data has already been useful in our disease studies and, upon completion, will be an invaluable resource for the interpretation of data derived from studies involving individuals of European descent. KIR haplotypic variability across ethnic groups appears to be extensive, as indicated by KIR gene profiles determined in unrelated individuals in a limited number of populations. These differences in KIR gene/allele frequencies may explain some of the observed variability in disease susceptibility across ethnic groups, particularly in regard to HLA class I associations. We have almost completed typing of samples from 29 populations for both KIR and HLA class I. Given the breadth of the population distributions throughout the world represented in this sample, we believe that this study will comprehensively define the extent of inter- and intra-population KIR haplotypic variability, potentially suggesting a role for selection in the generation of diversity across populations. Comparisons of individual KIR gene sequences within and across primate species indicate that the KIR gene family is characterized by historic instability, perhaps evolving rapidly in response to species-specific pathogenic organisms. KIR genes have now been identified in several primate species and KIR-like genes were identified in cow. In spite of the distinctions, some similarities have persisted over tens of millions of years, including the maintenance of KIR2DL4 in all primate species tested. While these studies have provided useful information regarding the evolution of KIR genes across primate species, comparisons of gene order and complete haplotypic sequence in different primate species would serve as a solid foundation for more accurately reconstructing the evolutionary history of the locus. Furthermore, a complete haplotypic KIR sequence in rhesus monkey, the species that serves as the primary animal model for human HIV disease, would greatly foster functional studies regarding the role of natural killer (NK) cell activity (as regulated by KIR molecules) in AIDS pathogenesis. We have screened rhesus monkey and chimpanzee bacterial artificial chromosome (BAC) libraries using a 287 bp probe representing exon 4 of KIR2DL4. Probes representing the NKp46 and ILT1 genes, which map on opposite sides of the KIR gene cluster in human, were used to identify BAC clones that cover the entire KIR complex. Eight clones from rhesus monkey and 12 clones from chimpanzee were identified that cross-hybridized with the KIR2DL4 probe and with either the NKp46 or ILT1 probes. In collaboration with Dr. Stephan Beck at the Sanger Center (Cambridge University, UK), the BAC clones were assembled into contigs, and three clones (two clones from chimpanzee and one from rhesus monkey) that were deemed likely to cover the full KIR region in each species were selected for sequencing. Dr. Beck's group has now completed sequencing of all three clones and annotation of the sequences has been performed. Recently, BAC libraries from a prosimian species, galago, and a new world monkey, the common marmoset, have become available through Dr. Pieter de Jong. We intend to screen these libraries for the KIR locus and Dr. Beck's group will sequence these clones, as well. A comparative genomics study will be performed in collaboration with Dr. John Trowsdale. KIR haplotypes generally range in gene content from about 6-12 expressed genes, although rare haplotypes that are very short (<6 genes) or very long (>12 genes) have been observed. The gene content and sequence of some KIR alleles strongly suggests that unequal crossing over in the region may account for much of the expansion and contraction this locus has undergone over time. For example, we have recently identified a hybrid gene of two distinct KIR genes, KIR2DL5A and KIR3DP1 on a haplotype that contains two copies of two different KIR genes. We have proposed a model involving unequal crossing over that could explain this haplotype. We have determined the genomic order of the KIR genes on this haplotype, supporting a mechanism involving unequal crossing over in the derivation of this haplotype. We hypothesize that nonreciprocal recombination within the KIR gene complex occurs as frequently (and perhaps more frequently) as reciprocal recombination due to the 16 kb repetitive segments characteristic of the KIR locus that may misalign during synapsis. We will use single sperm typing to rapidly screen for recombination within the KIR gene complex. The KIR locus is roughly 150 kb in length on average and the average male-specific recombination rate for chromosome 19 is 1.43 cM/Mb. If we assume a more conservative rate of 1 cM/Mb for the KIR gene cluster, then we would expect to observe about 15 recombinant chromosomes from a screen of 15,000 sperm using two informative markers on opposite ends of the KIR gene complex. DNA representing KIR locus recombinant chromosomes will be fully typed and compared to the KIR genotypes determined from the two nonrecombinant KIR haplotypes in order to determine whether unequal or equal crossing over accounts for the novel recombinant chromosome. This study will complement and directly address a possible mechanism for the expansion and contraction that has occurred within the KIR gene complex.